{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 336,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.008928571428571428,
      "grad_norm": 5.458911610311603,
      "learning_rate": 2.3529411764705885e-06,
      "loss": 0.9075,
      "step": 1
    },
    {
      "epoch": 0.017857142857142856,
      "grad_norm": 5.459968806476413,
      "learning_rate": 4.705882352941177e-06,
      "loss": 0.9041,
      "step": 2
    },
    {
      "epoch": 0.026785714285714284,
      "grad_norm": 5.116219510257077,
      "learning_rate": 7.058823529411766e-06,
      "loss": 0.8974,
      "step": 3
    },
    {
      "epoch": 0.03571428571428571,
      "grad_norm": 3.7608817755792816,
      "learning_rate": 9.411764705882354e-06,
      "loss": 0.8634,
      "step": 4
    },
    {
      "epoch": 0.044642857142857144,
      "grad_norm": 1.9769831033319734,
      "learning_rate": 1.1764705882352942e-05,
      "loss": 0.8282,
      "step": 5
    },
    {
      "epoch": 0.05357142857142857,
      "grad_norm": 5.097457631844493,
      "learning_rate": 1.4117647058823532e-05,
      "loss": 0.8469,
      "step": 6
    },
    {
      "epoch": 0.0625,
      "grad_norm": 6.677710780281878,
      "learning_rate": 1.647058823529412e-05,
      "loss": 0.8328,
      "step": 7
    },
    {
      "epoch": 0.07142857142857142,
      "grad_norm": 6.801682007638335,
      "learning_rate": 1.8823529411764708e-05,
      "loss": 0.8458,
      "step": 8
    },
    {
      "epoch": 0.08035714285714286,
      "grad_norm": 3.7515376420989193,
      "learning_rate": 2.1176470588235296e-05,
      "loss": 0.7872,
      "step": 9
    },
    {
      "epoch": 0.08928571428571429,
      "grad_norm": 2.8621020997538817,
      "learning_rate": 2.3529411764705884e-05,
      "loss": 0.7587,
      "step": 10
    },
    {
      "epoch": 0.09821428571428571,
      "grad_norm": 2.451873169585479,
      "learning_rate": 2.5882352941176475e-05,
      "loss": 0.7254,
      "step": 11
    },
    {
      "epoch": 0.10714285714285714,
      "grad_norm": 1.409713255422223,
      "learning_rate": 2.8235294117647063e-05,
      "loss": 0.7001,
      "step": 12
    },
    {
      "epoch": 0.11607142857142858,
      "grad_norm": 1.3093345645984427,
      "learning_rate": 3.0588235294117644e-05,
      "loss": 0.6808,
      "step": 13
    },
    {
      "epoch": 0.125,
      "grad_norm": 1.0595350381669622,
      "learning_rate": 3.294117647058824e-05,
      "loss": 0.6725,
      "step": 14
    },
    {
      "epoch": 0.13392857142857142,
      "grad_norm": 1.1345682356803515,
      "learning_rate": 3.529411764705883e-05,
      "loss": 0.6625,
      "step": 15
    },
    {
      "epoch": 0.14285714285714285,
      "grad_norm": 1.1508222129857923,
      "learning_rate": 3.7647058823529415e-05,
      "loss": 0.6508,
      "step": 16
    },
    {
      "epoch": 0.15178571428571427,
      "grad_norm": 1.3093404043153851,
      "learning_rate": 4e-05,
      "loss": 0.6344,
      "step": 17
    },
    {
      "epoch": 0.16071428571428573,
      "grad_norm": 0.9342090430783941,
      "learning_rate": 4.235294117647059e-05,
      "loss": 0.6347,
      "step": 18
    },
    {
      "epoch": 0.16964285714285715,
      "grad_norm": 1.1694538889722852,
      "learning_rate": 4.470588235294118e-05,
      "loss": 0.626,
      "step": 19
    },
    {
      "epoch": 0.17857142857142858,
      "grad_norm": 1.3017562899458819,
      "learning_rate": 4.705882352941177e-05,
      "loss": 0.6116,
      "step": 20
    },
    {
      "epoch": 0.1875,
      "grad_norm": 1.3227965952570264,
      "learning_rate": 4.941176470588236e-05,
      "loss": 0.6116,
      "step": 21
    },
    {
      "epoch": 0.19642857142857142,
      "grad_norm": 0.6130823943695296,
      "learning_rate": 5.176470588235295e-05,
      "loss": 0.607,
      "step": 22
    },
    {
      "epoch": 0.20535714285714285,
      "grad_norm": 1.251454832596592,
      "learning_rate": 5.411764705882354e-05,
      "loss": 0.6033,
      "step": 23
    },
    {
      "epoch": 0.21428571428571427,
      "grad_norm": 0.8156899349803868,
      "learning_rate": 5.6470588235294126e-05,
      "loss": 0.6052,
      "step": 24
    },
    {
      "epoch": 0.22321428571428573,
      "grad_norm": 1.37947321601158,
      "learning_rate": 5.8823529411764714e-05,
      "loss": 0.6063,
      "step": 25
    },
    {
      "epoch": 0.23214285714285715,
      "grad_norm": 1.0776438582531838,
      "learning_rate": 6.117647058823529e-05,
      "loss": 0.5931,
      "step": 26
    },
    {
      "epoch": 0.24107142857142858,
      "grad_norm": 1.3960841262426464,
      "learning_rate": 6.352941176470589e-05,
      "loss": 0.584,
      "step": 27
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.8924864474776818,
      "learning_rate": 6.588235294117648e-05,
      "loss": 0.5893,
      "step": 28
    },
    {
      "epoch": 0.25892857142857145,
      "grad_norm": 0.969188961164555,
      "learning_rate": 6.823529411764707e-05,
      "loss": 0.5852,
      "step": 29
    },
    {
      "epoch": 0.26785714285714285,
      "grad_norm": 1.3198098256340314,
      "learning_rate": 7.058823529411765e-05,
      "loss": 0.5904,
      "step": 30
    },
    {
      "epoch": 0.2767857142857143,
      "grad_norm": 1.4766678918535208,
      "learning_rate": 7.294117647058824e-05,
      "loss": 0.588,
      "step": 31
    },
    {
      "epoch": 0.2857142857142857,
      "grad_norm": 1.0600070593717186,
      "learning_rate": 7.529411764705883e-05,
      "loss": 0.5903,
      "step": 32
    },
    {
      "epoch": 0.29464285714285715,
      "grad_norm": 1.4709971965977178,
      "learning_rate": 7.764705882352942e-05,
      "loss": 0.5856,
      "step": 33
    },
    {
      "epoch": 0.30357142857142855,
      "grad_norm": 9.732884272846816,
      "learning_rate": 8e-05,
      "loss": 0.5782,
      "step": 34
    },
    {
      "epoch": 0.3125,
      "grad_norm": 1.471329125996165,
      "learning_rate": 7.999783572751229e-05,
      "loss": 0.591,
      "step": 35
    },
    {
      "epoch": 0.32142857142857145,
      "grad_norm": 0.9582405842286696,
      "learning_rate": 7.999134314425292e-05,
      "loss": 0.5654,
      "step": 36
    },
    {
      "epoch": 0.33035714285714285,
      "grad_norm": 1.3392752737259712,
      "learning_rate": 7.998052295280786e-05,
      "loss": 0.5629,
      "step": 37
    },
    {
      "epoch": 0.3392857142857143,
      "grad_norm": 1.0198091139639283,
      "learning_rate": 7.996537632406925e-05,
      "loss": 0.5701,
      "step": 38
    },
    {
      "epoch": 0.3482142857142857,
      "grad_norm": 0.9130649378368526,
      "learning_rate": 7.994590489710867e-05,
      "loss": 0.5667,
      "step": 39
    },
    {
      "epoch": 0.35714285714285715,
      "grad_norm": 1.1018536674455315,
      "learning_rate": 7.992211077899981e-05,
      "loss": 0.5621,
      "step": 40
    },
    {
      "epoch": 0.36607142857142855,
      "grad_norm": 1.3069162163076586,
      "learning_rate": 7.989399654459043e-05,
      "loss": 0.557,
      "step": 41
    },
    {
      "epoch": 0.375,
      "grad_norm": 0.9331234760948489,
      "learning_rate": 7.986156523622373e-05,
      "loss": 0.5621,
      "step": 42
    },
    {
      "epoch": 0.38392857142857145,
      "grad_norm": 0.72739280302451,
      "learning_rate": 7.982482036340912e-05,
      "loss": 0.5562,
      "step": 43
    },
    {
      "epoch": 0.39285714285714285,
      "grad_norm": 0.9580174839293588,
      "learning_rate": 7.978376590244249e-05,
      "loss": 0.5648,
      "step": 44
    },
    {
      "epoch": 0.4017857142857143,
      "grad_norm": 1.171231893769441,
      "learning_rate": 7.973840629597584e-05,
      "loss": 0.5549,
      "step": 45
    },
    {
      "epoch": 0.4107142857142857,
      "grad_norm": 0.6910616601450175,
      "learning_rate": 7.968874645253661e-05,
      "loss": 0.5554,
      "step": 46
    },
    {
      "epoch": 0.41964285714285715,
      "grad_norm": 0.6712508262420084,
      "learning_rate": 7.963479174599642e-05,
      "loss": 0.5539,
      "step": 47
    },
    {
      "epoch": 0.42857142857142855,
      "grad_norm": 0.8981493488966513,
      "learning_rate": 7.957654801498961e-05,
      "loss": 0.551,
      "step": 48
    },
    {
      "epoch": 0.4375,
      "grad_norm": 0.9743397904480438,
      "learning_rate": 7.951402156228145e-05,
      "loss": 0.5454,
      "step": 49
    },
    {
      "epoch": 0.44642857142857145,
      "grad_norm": 0.6631707178264241,
      "learning_rate": 7.944721915408596e-05,
      "loss": 0.534,
      "step": 50
    },
    {
      "epoch": 0.45535714285714285,
      "grad_norm": 0.6793775277783014,
      "learning_rate": 7.937614801933388e-05,
      "loss": 0.5451,
      "step": 51
    },
    {
      "epoch": 0.4642857142857143,
      "grad_norm": 0.8605556642034515,
      "learning_rate": 7.930081584889029e-05,
      "loss": 0.5403,
      "step": 52
    },
    {
      "epoch": 0.4732142857142857,
      "grad_norm": 0.7496935145980926,
      "learning_rate": 7.922123079472237e-05,
      "loss": 0.5437,
      "step": 53
    },
    {
      "epoch": 0.48214285714285715,
      "grad_norm": 0.5579609752590933,
      "learning_rate": 7.913740146901729e-05,
      "loss": 0.5352,
      "step": 54
    },
    {
      "epoch": 0.49107142857142855,
      "grad_norm": 0.7574194668910744,
      "learning_rate": 7.90493369432502e-05,
      "loss": 0.5402,
      "step": 55
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.9921442886534483,
      "learning_rate": 7.895704674720265e-05,
      "loss": 0.53,
      "step": 56
    },
    {
      "epoch": 0.5089285714285714,
      "grad_norm": 0.8243330121689626,
      "learning_rate": 7.88605408679312e-05,
      "loss": 0.5456,
      "step": 57
    },
    {
      "epoch": 0.5178571428571429,
      "grad_norm": 0.5792225695439277,
      "learning_rate": 7.875982974868684e-05,
      "loss": 0.5354,
      "step": 58
    },
    {
      "epoch": 0.5267857142857143,
      "grad_norm": 0.7457644638076486,
      "learning_rate": 7.865492428778483e-05,
      "loss": 0.5389,
      "step": 59
    },
    {
      "epoch": 0.5357142857142857,
      "grad_norm": 1.0227407044588295,
      "learning_rate": 7.854583583742527e-05,
      "loss": 0.5323,
      "step": 60
    },
    {
      "epoch": 0.5446428571428571,
      "grad_norm": 1.1493686113568151,
      "learning_rate": 7.843257620246477e-05,
      "loss": 0.5365,
      "step": 61
    },
    {
      "epoch": 0.5535714285714286,
      "grad_norm": 0.6003138642474546,
      "learning_rate": 7.831515763913891e-05,
      "loss": 0.5301,
      "step": 62
    },
    {
      "epoch": 0.5625,
      "grad_norm": 0.5676139643526338,
      "learning_rate": 7.819359285373604e-05,
      "loss": 0.5246,
      "step": 63
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 0.42378345040062193,
      "learning_rate": 7.806789500122212e-05,
      "loss": 0.5237,
      "step": 64
    },
    {
      "epoch": 0.5803571428571429,
      "grad_norm": 0.8692395127139948,
      "learning_rate": 7.793807768381743e-05,
      "loss": 0.52,
      "step": 65
    },
    {
      "epoch": 0.5892857142857143,
      "grad_norm": 0.7123093455414804,
      "learning_rate": 7.780415494952431e-05,
      "loss": 0.5313,
      "step": 66
    },
    {
      "epoch": 0.5982142857142857,
      "grad_norm": 0.7693658767139915,
      "learning_rate": 7.766614129060728e-05,
      "loss": 0.5206,
      "step": 67
    },
    {
      "epoch": 0.6071428571428571,
      "grad_norm": 0.5764493737404597,
      "learning_rate": 7.752405164202459e-05,
      "loss": 0.531,
      "step": 68
    },
    {
      "epoch": 0.6160714285714286,
      "grad_norm": 2.718091673505057,
      "learning_rate": 7.737790137981205e-05,
      "loss": 0.5296,
      "step": 69
    },
    {
      "epoch": 0.625,
      "grad_norm": 0.8095852508927421,
      "learning_rate": 7.722770631941928e-05,
      "loss": 0.5288,
      "step": 70
    },
    {
      "epoch": 0.6339285714285714,
      "grad_norm": 3.3841935199500925,
      "learning_rate": 7.707348271399813e-05,
      "loss": 0.5502,
      "step": 71
    },
    {
      "epoch": 0.6428571428571429,
      "grad_norm": 1.745098535789061,
      "learning_rate": 7.691524725264389e-05,
      "loss": 0.5516,
      "step": 72
    },
    {
      "epoch": 0.6517857142857143,
      "grad_norm": 0.5725808597680306,
      "learning_rate": 7.675301705858935e-05,
      "loss": 0.5363,
      "step": 73
    },
    {
      "epoch": 0.6607142857142857,
      "grad_norm": 1.2584222907391829,
      "learning_rate": 7.658680968735181e-05,
      "loss": 0.5402,
      "step": 74
    },
    {
      "epoch": 0.6696428571428571,
      "grad_norm": 0.6752011986304464,
      "learning_rate": 7.641664312483328e-05,
      "loss": 0.5321,
      "step": 75
    },
    {
      "epoch": 0.6785714285714286,
      "grad_norm": 0.9309963276595731,
      "learning_rate": 7.624253578537424e-05,
      "loss": 0.532,
      "step": 76
    },
    {
      "epoch": 0.6875,
      "grad_norm": 0.6408984967666898,
      "learning_rate": 7.606450650976095e-05,
      "loss": 0.5367,
      "step": 77
    },
    {
      "epoch": 0.6964285714285714,
      "grad_norm": 0.6349526215998039,
      "learning_rate": 7.588257456318657e-05,
      "loss": 0.5302,
      "step": 78
    },
    {
      "epoch": 0.7053571428571429,
      "grad_norm": 0.5614114341482542,
      "learning_rate": 7.56967596331664e-05,
      "loss": 0.5202,
      "step": 79
    },
    {
      "epoch": 0.7142857142857143,
      "grad_norm": 0.5310297157289734,
      "learning_rate": 7.550708182740751e-05,
      "loss": 0.5327,
      "step": 80
    },
    {
      "epoch": 0.7232142857142857,
      "grad_norm": 0.4999950300313544,
      "learning_rate": 7.531356167163272e-05,
      "loss": 0.5256,
      "step": 81
    },
    {
      "epoch": 0.7321428571428571,
      "grad_norm": 0.41402163485309396,
      "learning_rate": 7.511622010735948e-05,
      "loss": 0.5241,
      "step": 82
    },
    {
      "epoch": 0.7410714285714286,
      "grad_norm": 0.43962685575693505,
      "learning_rate": 7.49150784896337e-05,
      "loss": 0.5215,
      "step": 83
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.3656133936292797,
      "learning_rate": 7.471015858471882e-05,
      "loss": 0.5081,
      "step": 84
    },
    {
      "epoch": 0.7589285714285714,
      "grad_norm": 0.3744198516273023,
      "learning_rate": 7.450148256774053e-05,
      "loss": 0.5142,
      "step": 85
    },
    {
      "epoch": 0.7678571428571429,
      "grad_norm": 0.40929219272400663,
      "learning_rate": 7.428907302028687e-05,
      "loss": 0.5196,
      "step": 86
    },
    {
      "epoch": 0.7767857142857143,
      "grad_norm": 0.4065715708969531,
      "learning_rate": 7.407295292796488e-05,
      "loss": 0.5204,
      "step": 87
    },
    {
      "epoch": 0.7857142857142857,
      "grad_norm": 0.34204445173011166,
      "learning_rate": 7.385314567791302e-05,
      "loss": 0.5183,
      "step": 88
    },
    {
      "epoch": 0.7946428571428571,
      "grad_norm": 0.32805261246572137,
      "learning_rate": 7.362967505627051e-05,
      "loss": 0.5213,
      "step": 89
    },
    {
      "epoch": 0.8035714285714286,
      "grad_norm": 0.3272013216504384,
      "learning_rate": 7.340256524560325e-05,
      "loss": 0.512,
      "step": 90
    },
    {
      "epoch": 0.8125,
      "grad_norm": 0.293354381788003,
      "learning_rate": 7.317184082228697e-05,
      "loss": 0.5069,
      "step": 91
    },
    {
      "epoch": 0.8214285714285714,
      "grad_norm": 0.4821061440980201,
      "learning_rate": 7.293752675384777e-05,
      "loss": 0.5107,
      "step": 92
    },
    {
      "epoch": 0.8303571428571429,
      "grad_norm": 0.693462071819636,
      "learning_rate": 7.269964839626024e-05,
      "loss": 0.5184,
      "step": 93
    },
    {
      "epoch": 0.8392857142857143,
      "grad_norm": 0.9969190523365147,
      "learning_rate": 7.245823149120362e-05,
      "loss": 0.5249,
      "step": 94
    },
    {
      "epoch": 0.8482142857142857,
      "grad_norm": 0.9850693615579191,
      "learning_rate": 7.221330216327619e-05,
      "loss": 0.5218,
      "step": 95
    },
    {
      "epoch": 0.8571428571428571,
      "grad_norm": 0.7805590557073407,
      "learning_rate": 7.196488691716824e-05,
      "loss": 0.5152,
      "step": 96
    },
    {
      "epoch": 0.8660714285714286,
      "grad_norm": 0.5570109379564657,
      "learning_rate": 7.17130126347939e-05,
      "loss": 0.5118,
      "step": 97
    },
    {
      "epoch": 0.875,
      "grad_norm": 0.5791838534781877,
      "learning_rate": 7.145770657238215e-05,
      "loss": 0.5095,
      "step": 98
    },
    {
      "epoch": 0.8839285714285714,
      "grad_norm": 0.6535836061766679,
      "learning_rate": 7.119899635752737e-05,
      "loss": 0.5066,
      "step": 99
    },
    {
      "epoch": 0.8928571428571429,
      "grad_norm": 0.4994295705575278,
      "learning_rate": 7.093690998619953e-05,
      "loss": 0.5108,
      "step": 100
    },
    {
      "epoch": 0.9017857142857143,
      "grad_norm": 0.5215314045730167,
      "learning_rate": 7.067147581971478e-05,
      "loss": 0.5144,
      "step": 101
    },
    {
      "epoch": 0.9107142857142857,
      "grad_norm": 0.556364903677944,
      "learning_rate": 7.040272258166633e-05,
      "loss": 0.5088,
      "step": 102
    },
    {
      "epoch": 0.9196428571428571,
      "grad_norm": 0.3522985441955947,
      "learning_rate": 7.013067935481614e-05,
      "loss": 0.5236,
      "step": 103
    },
    {
      "epoch": 0.9285714285714286,
      "grad_norm": 0.5091651011552728,
      "learning_rate": 6.985537557794776e-05,
      "loss": 0.501,
      "step": 104
    },
    {
      "epoch": 0.9375,
      "grad_norm": 0.3561680464897058,
      "learning_rate": 6.957684104268068e-05,
      "loss": 0.5023,
      "step": 105
    },
    {
      "epoch": 0.9464285714285714,
      "grad_norm": 0.40001485875356707,
      "learning_rate": 6.92951058902465e-05,
      "loss": 0.5066,
      "step": 106
    },
    {
      "epoch": 0.9553571428571429,
      "grad_norm": 0.36789962565963513,
      "learning_rate": 6.901020060822718e-05,
      "loss": 0.5042,
      "step": 107
    },
    {
      "epoch": 0.9642857142857143,
      "grad_norm": 0.4184528762777128,
      "learning_rate": 6.872215602725589e-05,
      "loss": 0.5128,
      "step": 108
    },
    {
      "epoch": 0.9732142857142857,
      "grad_norm": 0.4469489989551305,
      "learning_rate": 6.843100331768071e-05,
      "loss": 0.5042,
      "step": 109
    },
    {
      "epoch": 0.9821428571428571,
      "grad_norm": 0.3927827934106694,
      "learning_rate": 6.813677398619161e-05,
      "loss": 0.5014,
      "step": 110
    },
    {
      "epoch": 0.9910714285714286,
      "grad_norm": 0.30551724357758264,
      "learning_rate": 6.783949987241095e-05,
      "loss": 0.4939,
      "step": 111
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.3878355524761787,
      "learning_rate": 6.7539213145448e-05,
      "loss": 0.5016,
      "step": 112
    },
    {
      "epoch": 1.0089285714285714,
      "grad_norm": 0.3788672363521083,
      "learning_rate": 6.723594630041786e-05,
      "loss": 0.4729,
      "step": 113
    },
    {
      "epoch": 1.0178571428571428,
      "grad_norm": 0.3055300238162846,
      "learning_rate": 6.692973215492498e-05,
      "loss": 0.4692,
      "step": 114
    },
    {
      "epoch": 1.0267857142857142,
      "grad_norm": 0.3859310266273548,
      "learning_rate": 6.662060384551188e-05,
      "loss": 0.4707,
      "step": 115
    },
    {
      "epoch": 1.0357142857142858,
      "grad_norm": 0.5582243479324142,
      "learning_rate": 6.630859482407331e-05,
      "loss": 0.4722,
      "step": 116
    },
    {
      "epoch": 1.0446428571428572,
      "grad_norm": 0.6769506937788301,
      "learning_rate": 6.599373885423635e-05,
      "loss": 0.4729,
      "step": 117
    },
    {
      "epoch": 1.0535714285714286,
      "grad_norm": 0.6799095444990083,
      "learning_rate": 6.567607000770664e-05,
      "loss": 0.4778,
      "step": 118
    },
    {
      "epoch": 1.0625,
      "grad_norm": 0.6711015115485335,
      "learning_rate": 6.535562266058141e-05,
      "loss": 0.4783,
      "step": 119
    },
    {
      "epoch": 1.0714285714285714,
      "grad_norm": 0.6616384329423337,
      "learning_rate": 6.50324314896295e-05,
      "loss": 0.483,
      "step": 120
    },
    {
      "epoch": 1.0803571428571428,
      "grad_norm": 0.5335329011878119,
      "learning_rate": 6.470653146853895e-05,
      "loss": 0.4676,
      "step": 121
    },
    {
      "epoch": 1.0892857142857142,
      "grad_norm": 0.5280180658388947,
      "learning_rate": 6.437795786413217e-05,
      "loss": 0.4676,
      "step": 122
    },
    {
      "epoch": 1.0982142857142858,
      "grad_norm": 0.5203647507319533,
      "learning_rate": 6.40467462325498e-05,
      "loss": 0.4767,
      "step": 123
    },
    {
      "epoch": 1.1071428571428572,
      "grad_norm": 0.33866738617528636,
      "learning_rate": 6.371293241540292e-05,
      "loss": 0.4649,
      "step": 124
    },
    {
      "epoch": 1.1160714285714286,
      "grad_norm": 0.39685847712765815,
      "learning_rate": 6.337655253589456e-05,
      "loss": 0.4654,
      "step": 125
    },
    {
      "epoch": 1.125,
      "grad_norm": 0.2986911683728224,
      "learning_rate": 6.303764299491065e-05,
      "loss": 0.4656,
      "step": 126
    },
    {
      "epoch": 1.1339285714285714,
      "grad_norm": 0.27795327083232535,
      "learning_rate": 6.269624046708094e-05,
      "loss": 0.4674,
      "step": 127
    },
    {
      "epoch": 1.1428571428571428,
      "grad_norm": 0.2582126483978543,
      "learning_rate": 6.235238189681039e-05,
      "loss": 0.4653,
      "step": 128
    },
    {
      "epoch": 1.1517857142857142,
      "grad_norm": 0.26393000429448243,
      "learning_rate": 6.200610449428111e-05,
      "loss": 0.4643,
      "step": 129
    },
    {
      "epoch": 1.1607142857142858,
      "grad_norm": 0.2717121850307768,
      "learning_rate": 6.16574457314259e-05,
      "loss": 0.475,
      "step": 130
    },
    {
      "epoch": 1.1696428571428572,
      "grad_norm": 0.29888138759375577,
      "learning_rate": 6.130644333787317e-05,
      "loss": 0.4627,
      "step": 131
    },
    {
      "epoch": 1.1785714285714286,
      "grad_norm": 0.32040905560153704,
      "learning_rate": 6.095313529686406e-05,
      "loss": 0.4609,
      "step": 132
    },
    {
      "epoch": 1.1875,
      "grad_norm": 0.31544315623213,
      "learning_rate": 6.059755984114224e-05,
      "loss": 0.4591,
      "step": 133
    },
    {
      "epoch": 1.1964285714285714,
      "grad_norm": 0.35573100943666236,
      "learning_rate": 6.02397554488165e-05,
      "loss": 0.4655,
      "step": 134
    },
    {
      "epoch": 1.2053571428571428,
      "grad_norm": 0.3371073630997141,
      "learning_rate": 5.987976083919699e-05,
      "loss": 0.4602,
      "step": 135
    },
    {
      "epoch": 1.2142857142857142,
      "grad_norm": 0.24367143463905358,
      "learning_rate": 5.951761496860513e-05,
      "loss": 0.4686,
      "step": 136
    },
    {
      "epoch": 1.2232142857142858,
      "grad_norm": 0.18490249219086996,
      "learning_rate": 5.915335702615817e-05,
      "loss": 0.4694,
      "step": 137
    },
    {
      "epoch": 1.2321428571428572,
      "grad_norm": 0.3662203733341645,
      "learning_rate": 5.878702642952824e-05,
      "loss": 0.4562,
      "step": 138
    },
    {
      "epoch": 1.2410714285714286,
      "grad_norm": 0.44890191448322514,
      "learning_rate": 5.8418662820676954e-05,
      "loss": 0.4639,
      "step": 139
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.4261072543726084,
      "learning_rate": 5.80483060615655e-05,
      "loss": 0.4732,
      "step": 140
    },
    {
      "epoch": 1.2589285714285714,
      "grad_norm": 0.3422728954134477,
      "learning_rate": 5.7675996229841107e-05,
      "loss": 0.4619,
      "step": 141
    },
    {
      "epoch": 1.2678571428571428,
      "grad_norm": 0.3076389147366431,
      "learning_rate": 5.730177361450005e-05,
      "loss": 0.4722,
      "step": 142
    },
    {
      "epoch": 1.2767857142857144,
      "grad_norm": 0.2820370998236569,
      "learning_rate": 5.6925678711527865e-05,
      "loss": 0.468,
      "step": 143
    },
    {
      "epoch": 1.2857142857142856,
      "grad_norm": 0.21191838816193023,
      "learning_rate": 5.6547752219517136e-05,
      "loss": 0.4595,
      "step": 144
    },
    {
      "epoch": 1.2946428571428572,
      "grad_norm": 0.18589392586021203,
      "learning_rate": 5.616803503526328e-05,
      "loss": 0.4644,
      "step": 145
    },
    {
      "epoch": 1.3035714285714286,
      "grad_norm": 0.20528736580535784,
      "learning_rate": 5.578656824933906e-05,
      "loss": 0.458,
      "step": 146
    },
    {
      "epoch": 1.3125,
      "grad_norm": 0.25935591546202547,
      "learning_rate": 5.540339314164798e-05,
      "loss": 0.4625,
      "step": 147
    },
    {
      "epoch": 1.3214285714285714,
      "grad_norm": 0.2603875779161184,
      "learning_rate": 5.501855117695721e-05,
      "loss": 0.464,
      "step": 148
    },
    {
      "epoch": 1.3303571428571428,
      "grad_norm": 0.259347292325694,
      "learning_rate": 5.463208400041055e-05,
      "loss": 0.4697,
      "step": 149
    },
    {
      "epoch": 1.3392857142857144,
      "grad_norm": 0.2644029291060694,
      "learning_rate": 5.424403343302188e-05,
      "loss": 0.4576,
      "step": 150
    },
    {
      "epoch": 1.3482142857142856,
      "grad_norm": 0.2308783303320916,
      "learning_rate": 5.385444146714957e-05,
      "loss": 0.4564,
      "step": 151
    },
    {
      "epoch": 1.3571428571428572,
      "grad_norm": 0.22882568254344396,
      "learning_rate": 5.346335026195226e-05,
      "loss": 0.4721,
      "step": 152
    },
    {
      "epoch": 1.3660714285714286,
      "grad_norm": 0.23616899606363104,
      "learning_rate": 5.307080213882672e-05,
      "loss": 0.4662,
      "step": 153
    },
    {
      "epoch": 1.375,
      "grad_norm": 0.2885371827319689,
      "learning_rate": 5.267683957682812e-05,
      "loss": 0.462,
      "step": 154
    },
    {
      "epoch": 1.3839285714285714,
      "grad_norm": 0.2820779719230081,
      "learning_rate": 5.2281505208073147e-05,
      "loss": 0.4632,
      "step": 155
    },
    {
      "epoch": 1.3928571428571428,
      "grad_norm": 0.22468099441098213,
      "learning_rate": 5.18848418131267e-05,
      "loss": 0.4551,
      "step": 156
    },
    {
      "epoch": 1.4017857142857144,
      "grad_norm": 0.21751416007716054,
      "learning_rate": 5.1486892316372404e-05,
      "loss": 0.4722,
      "step": 157
    },
    {
      "epoch": 1.4107142857142856,
      "grad_norm": 0.27461792920575495,
      "learning_rate": 5.1087699781367625e-05,
      "loss": 0.463,
      "step": 158
    },
    {
      "epoch": 1.4196428571428572,
      "grad_norm": 0.2393265717043113,
      "learning_rate": 5.0687307406183406e-05,
      "loss": 0.4566,
      "step": 159
    },
    {
      "epoch": 1.4285714285714286,
      "grad_norm": 0.20221953908786175,
      "learning_rate": 5.028575851872984e-05,
      "loss": 0.4646,
      "step": 160
    },
    {
      "epoch": 1.4375,
      "grad_norm": 0.2208471324202507,
      "learning_rate": 4.988309657206742e-05,
      "loss": 0.4599,
      "step": 161
    },
    {
      "epoch": 1.4464285714285714,
      "grad_norm": 0.19928106325960415,
      "learning_rate": 4.947936513970476e-05,
      "loss": 0.4587,
      "step": 162
    },
    {
      "epoch": 1.4553571428571428,
      "grad_norm": 0.16352276342020416,
      "learning_rate": 4.907460791088347e-05,
      "loss": 0.4563,
      "step": 163
    },
    {
      "epoch": 1.4642857142857144,
      "grad_norm": 0.22322494415515243,
      "learning_rate": 4.8668868685850265e-05,
      "loss": 0.4607,
      "step": 164
    },
    {
      "epoch": 1.4732142857142856,
      "grad_norm": 0.28312396593030925,
      "learning_rate": 4.826219137111724e-05,
      "loss": 0.4596,
      "step": 165
    },
    {
      "epoch": 1.4821428571428572,
      "grad_norm": 0.20863485843019422,
      "learning_rate": 4.785461997471058e-05,
      "loss": 0.4572,
      "step": 166
    },
    {
      "epoch": 1.4910714285714286,
      "grad_norm": 0.18263430431801883,
      "learning_rate": 4.7446198601408274e-05,
      "loss": 0.461,
      "step": 167
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.21990339290691535,
      "learning_rate": 4.703697144796742e-05,
      "loss": 0.4546,
      "step": 168
    },
    {
      "epoch": 1.5089285714285714,
      "grad_norm": 0.1919102767561462,
      "learning_rate": 4.662698279834149e-05,
      "loss": 0.4592,
      "step": 169
    },
    {
      "epoch": 1.5178571428571428,
      "grad_norm": 0.17372269475870397,
      "learning_rate": 4.62162770188882e-05,
      "loss": 0.469,
      "step": 170
    },
    {
      "epoch": 1.5267857142857144,
      "grad_norm": 0.1654682068782294,
      "learning_rate": 4.580489855356852e-05,
      "loss": 0.4559,
      "step": 171
    },
    {
      "epoch": 1.5357142857142856,
      "grad_norm": 0.17629220459825287,
      "learning_rate": 4.5392891919137155e-05,
      "loss": 0.4561,
      "step": 172
    },
    {
      "epoch": 1.5446428571428572,
      "grad_norm": 0.21172057493487084,
      "learning_rate": 4.498030170032531e-05,
      "loss": 0.461,
      "step": 173
    },
    {
      "epoch": 1.5535714285714286,
      "grad_norm": 0.17570649332026045,
      "learning_rate": 4.456717254501595e-05,
      "loss": 0.461,
      "step": 174
    },
    {
      "epoch": 1.5625,
      "grad_norm": 0.16836022815630258,
      "learning_rate": 4.4153549159412276e-05,
      "loss": 0.4576,
      "step": 175
    },
    {
      "epoch": 1.5714285714285714,
      "grad_norm": 0.1749663199663638,
      "learning_rate": 4.37394763032e-05,
      "loss": 0.4549,
      "step": 176
    },
    {
      "epoch": 1.5803571428571428,
      "grad_norm": 0.17474917301910808,
      "learning_rate": 4.332499878470367e-05,
      "loss": 0.4631,
      "step": 177
    },
    {
      "epoch": 1.5892857142857144,
      "grad_norm": 0.14004770259306568,
      "learning_rate": 4.291016145603776e-05,
      "loss": 0.4562,
      "step": 178
    },
    {
      "epoch": 1.5982142857142856,
      "grad_norm": 0.19029016036379953,
      "learning_rate": 4.249500920825314e-05,
      "loss": 0.4616,
      "step": 179
    },
    {
      "epoch": 1.6071428571428572,
      "grad_norm": 0.19188224946309962,
      "learning_rate": 4.2079586966479214e-05,
      "loss": 0.4636,
      "step": 180
    },
    {
      "epoch": 1.6160714285714286,
      "grad_norm": 0.18536862750852307,
      "learning_rate": 4.166393968506244e-05,
      "loss": 0.4596,
      "step": 181
    },
    {
      "epoch": 1.625,
      "grad_norm": 0.17003638172457017,
      "learning_rate": 4.124811234270157e-05,
      "loss": 0.4546,
      "step": 182
    },
    {
      "epoch": 1.6339285714285714,
      "grad_norm": 0.18025420888895052,
      "learning_rate": 4.0832149937580457e-05,
      "loss": 0.4583,
      "step": 183
    },
    {
      "epoch": 1.6428571428571428,
      "grad_norm": 0.17702049953400414,
      "learning_rate": 4.0416097482498575e-05,
      "loss": 0.4583,
      "step": 184
    },
    {
      "epoch": 1.6517857142857144,
      "grad_norm": 0.16060960803693247,
      "learning_rate": 4e-05,
      "loss": 0.458,
      "step": 185
    },
    {
      "epoch": 1.6607142857142856,
      "grad_norm": 0.17703603102767604,
      "learning_rate": 3.958390251750144e-05,
      "loss": 0.4633,
      "step": 186
    },
    {
      "epoch": 1.6696428571428572,
      "grad_norm": 0.1646046678567389,
      "learning_rate": 3.916785006241955e-05,
      "loss": 0.4626,
      "step": 187
    },
    {
      "epoch": 1.6785714285714286,
      "grad_norm": 0.1722881349469007,
      "learning_rate": 3.875188765729843e-05,
      "loss": 0.4567,
      "step": 188
    },
    {
      "epoch": 1.6875,
      "grad_norm": 0.14735936478640962,
      "learning_rate": 3.833606031493758e-05,
      "loss": 0.4604,
      "step": 189
    },
    {
      "epoch": 1.6964285714285714,
      "grad_norm": 0.14292807283154454,
      "learning_rate": 3.792041303352079e-05,
      "loss": 0.4583,
      "step": 190
    },
    {
      "epoch": 1.7053571428571428,
      "grad_norm": 0.16040731375651313,
      "learning_rate": 3.7504990791746877e-05,
      "loss": 0.4551,
      "step": 191
    },
    {
      "epoch": 1.7142857142857144,
      "grad_norm": 0.12644736716718807,
      "learning_rate": 3.708983854396225e-05,
      "loss": 0.4596,
      "step": 192
    },
    {
      "epoch": 1.7232142857142856,
      "grad_norm": 0.15010655874057,
      "learning_rate": 3.667500121529634e-05,
      "loss": 0.4523,
      "step": 193
    },
    {
      "epoch": 1.7321428571428572,
      "grad_norm": 0.1433646952462073,
      "learning_rate": 3.626052369679999e-05,
      "loss": 0.4598,
      "step": 194
    },
    {
      "epoch": 1.7410714285714286,
      "grad_norm": 0.11933250859224095,
      "learning_rate": 3.584645084058774e-05,
      "loss": 0.4526,
      "step": 195
    },
    {
      "epoch": 1.75,
      "grad_norm": 0.14269745540138357,
      "learning_rate": 3.543282745498407e-05,
      "loss": 0.4609,
      "step": 196
    },
    {
      "epoch": 1.7589285714285714,
      "grad_norm": 0.13719631088911224,
      "learning_rate": 3.50196982996747e-05,
      "loss": 0.4569,
      "step": 197
    },
    {
      "epoch": 1.7678571428571428,
      "grad_norm": 0.14123727092169555,
      "learning_rate": 3.460710808086285e-05,
      "loss": 0.4598,
      "step": 198
    },
    {
      "epoch": 1.7767857142857144,
      "grad_norm": 0.13589211534733778,
      "learning_rate": 3.419510144643149e-05,
      "loss": 0.4583,
      "step": 199
    },
    {
      "epoch": 1.7857142857142856,
      "grad_norm": 0.14463228336420042,
      "learning_rate": 3.378372298111181e-05,
      "loss": 0.4547,
      "step": 200
    },
    {
      "epoch": 1.7946428571428572,
      "grad_norm": 0.13381794986766524,
      "learning_rate": 3.3373017201658514e-05,
      "loss": 0.4588,
      "step": 201
    },
    {
      "epoch": 1.8035714285714286,
      "grad_norm": 0.13115685617204254,
      "learning_rate": 3.2963028552032594e-05,
      "loss": 0.4566,
      "step": 202
    },
    {
      "epoch": 1.8125,
      "grad_norm": 0.13196620026552933,
      "learning_rate": 3.2553801398591746e-05,
      "loss": 0.4557,
      "step": 203
    },
    {
      "epoch": 1.8214285714285714,
      "grad_norm": 0.12289630749221309,
      "learning_rate": 3.214538002528943e-05,
      "loss": 0.4597,
      "step": 204
    },
    {
      "epoch": 1.8303571428571428,
      "grad_norm": 0.12502731241075712,
      "learning_rate": 3.173780862888277e-05,
      "loss": 0.4567,
      "step": 205
    },
    {
      "epoch": 1.8392857142857144,
      "grad_norm": 0.11886410880699969,
      "learning_rate": 3.133113131414975e-05,
      "loss": 0.4628,
      "step": 206
    },
    {
      "epoch": 1.8482142857142856,
      "grad_norm": 0.12359267043234633,
      "learning_rate": 3.092539208911654e-05,
      "loss": 0.4582,
      "step": 207
    },
    {
      "epoch": 1.8571428571428572,
      "grad_norm": 0.13017549097611902,
      "learning_rate": 3.0520634860295254e-05,
      "loss": 0.4521,
      "step": 208
    },
    {
      "epoch": 1.8660714285714286,
      "grad_norm": 0.11869071958078599,
      "learning_rate": 3.0116903427932603e-05,
      "loss": 0.4536,
      "step": 209
    },
    {
      "epoch": 1.875,
      "grad_norm": 0.13940208007631327,
      "learning_rate": 2.9714241481270168e-05,
      "loss": 0.4489,
      "step": 210
    },
    {
      "epoch": 1.8839285714285714,
      "grad_norm": 0.11974765997048924,
      "learning_rate": 2.9312692593816604e-05,
      "loss": 0.4487,
      "step": 211
    },
    {
      "epoch": 1.8928571428571428,
      "grad_norm": 0.12948893942349837,
      "learning_rate": 2.8912300218632385e-05,
      "loss": 0.4582,
      "step": 212
    },
    {
      "epoch": 1.9017857142857144,
      "grad_norm": 0.12275899875447714,
      "learning_rate": 2.8513107683627602e-05,
      "loss": 0.4662,
      "step": 213
    },
    {
      "epoch": 1.9107142857142856,
      "grad_norm": 0.12457402598657341,
      "learning_rate": 2.8115158186873307e-05,
      "loss": 0.4498,
      "step": 214
    },
    {
      "epoch": 1.9196428571428572,
      "grad_norm": 0.13773022448582894,
      "learning_rate": 2.771849479192687e-05,
      "loss": 0.4638,
      "step": 215
    },
    {
      "epoch": 1.9285714285714286,
      "grad_norm": 0.1382501463811384,
      "learning_rate": 2.7323160423171894e-05,
      "loss": 0.4587,
      "step": 216
    },
    {
      "epoch": 1.9375,
      "grad_norm": 0.13537863495707703,
      "learning_rate": 2.692919786117329e-05,
      "loss": 0.4536,
      "step": 217
    },
    {
      "epoch": 1.9464285714285714,
      "grad_norm": 0.1455865352799823,
      "learning_rate": 2.6536649738047756e-05,
      "loss": 0.4529,
      "step": 218
    },
    {
      "epoch": 1.9553571428571428,
      "grad_norm": 0.12729730828758748,
      "learning_rate": 2.614555853285044e-05,
      "loss": 0.4512,
      "step": 219
    },
    {
      "epoch": 1.9642857142857144,
      "grad_norm": 0.12357176809760773,
      "learning_rate": 2.5755966566978124e-05,
      "loss": 0.4549,
      "step": 220
    },
    {
      "epoch": 1.9732142857142856,
      "grad_norm": 0.1192751623548519,
      "learning_rate": 2.5367915999589462e-05,
      "loss": 0.4469,
      "step": 221
    },
    {
      "epoch": 1.9821428571428572,
      "grad_norm": 0.12358485840017747,
      "learning_rate": 2.498144882304281e-05,
      "loss": 0.4506,
      "step": 222
    },
    {
      "epoch": 1.9910714285714286,
      "grad_norm": 0.11913180318943509,
      "learning_rate": 2.4596606858352036e-05,
      "loss": 0.4574,
      "step": 223
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.12792742986464758,
      "learning_rate": 2.421343175066095e-05,
      "loss": 0.4567,
      "step": 224
    },
    {
      "epoch": 2.0089285714285716,
      "grad_norm": 0.18610386161673095,
      "learning_rate": 2.3831964964736737e-05,
      "loss": 0.4285,
      "step": 225
    },
    {
      "epoch": 2.017857142857143,
      "grad_norm": 0.1317157519843888,
      "learning_rate": 2.3452247780482877e-05,
      "loss": 0.4245,
      "step": 226
    },
    {
      "epoch": 2.0267857142857144,
      "grad_norm": 0.15732732092952154,
      "learning_rate": 2.3074321288472135e-05,
      "loss": 0.4143,
      "step": 227
    },
    {
      "epoch": 2.0357142857142856,
      "grad_norm": 0.16814184352306275,
      "learning_rate": 2.269822638549997e-05,
      "loss": 0.4182,
      "step": 228
    },
    {
      "epoch": 2.044642857142857,
      "grad_norm": 0.158387600350679,
      "learning_rate": 2.2324003770158917e-05,
      "loss": 0.4266,
      "step": 229
    },
    {
      "epoch": 2.0535714285714284,
      "grad_norm": 0.16005630429705614,
      "learning_rate": 2.1951693938434517e-05,
      "loss": 0.4163,
      "step": 230
    },
    {
      "epoch": 2.0625,
      "grad_norm": 0.1541823036416597,
      "learning_rate": 2.1581337179323062e-05,
      "loss": 0.4202,
      "step": 231
    },
    {
      "epoch": 2.0714285714285716,
      "grad_norm": 0.15428909223565776,
      "learning_rate": 2.1212973570471758e-05,
      "loss": 0.4204,
      "step": 232
    },
    {
      "epoch": 2.080357142857143,
      "grad_norm": 0.13935809750720052,
      "learning_rate": 2.0846642973841833e-05,
      "loss": 0.4255,
      "step": 233
    },
    {
      "epoch": 2.0892857142857144,
      "grad_norm": 0.1481665794140764,
      "learning_rate": 2.0482385031394864e-05,
      "loss": 0.4173,
      "step": 234
    },
    {
      "epoch": 2.0982142857142856,
      "grad_norm": 0.14270794298166453,
      "learning_rate": 2.0120239160803026e-05,
      "loss": 0.4253,
      "step": 235
    },
    {
      "epoch": 2.107142857142857,
      "grad_norm": 0.12890317989622468,
      "learning_rate": 1.9760244551183503e-05,
      "loss": 0.4234,
      "step": 236
    },
    {
      "epoch": 2.1160714285714284,
      "grad_norm": 0.14547199967318059,
      "learning_rate": 1.9402440158857775e-05,
      "loss": 0.4241,
      "step": 237
    },
    {
      "epoch": 2.125,
      "grad_norm": 0.13117058221551006,
      "learning_rate": 1.9046864703135958e-05,
      "loss": 0.4174,
      "step": 238
    },
    {
      "epoch": 2.1339285714285716,
      "grad_norm": 0.13210270356792833,
      "learning_rate": 1.8693556662126857e-05,
      "loss": 0.4211,
      "step": 239
    },
    {
      "epoch": 2.142857142857143,
      "grad_norm": 0.1282730434203236,
      "learning_rate": 1.8342554268574114e-05,
      "loss": 0.4158,
      "step": 240
    },
    {
      "epoch": 2.1517857142857144,
      "grad_norm": 0.12878778566197982,
      "learning_rate": 1.7993895505718907e-05,
      "loss": 0.4164,
      "step": 241
    },
    {
      "epoch": 2.1607142857142856,
      "grad_norm": 0.12711242314558882,
      "learning_rate": 1.7647618103189632e-05,
      "loss": 0.4213,
      "step": 242
    },
    {
      "epoch": 2.169642857142857,
      "grad_norm": 0.12225869755757035,
      "learning_rate": 1.7303759532919066e-05,
      "loss": 0.4194,
      "step": 243
    },
    {
      "epoch": 2.1785714285714284,
      "grad_norm": 0.12636232755876087,
      "learning_rate": 1.6962357005089368e-05,
      "loss": 0.4171,
      "step": 244
    },
    {
      "epoch": 2.1875,
      "grad_norm": 0.11732096389813856,
      "learning_rate": 1.6623447464105455e-05,
      "loss": 0.4218,
      "step": 245
    },
    {
      "epoch": 2.1964285714285716,
      "grad_norm": 0.12702977153503237,
      "learning_rate": 1.628706758459709e-05,
      "loss": 0.4184,
      "step": 246
    },
    {
      "epoch": 2.205357142857143,
      "grad_norm": 0.11036816052659255,
      "learning_rate": 1.5953253767450214e-05,
      "loss": 0.4233,
      "step": 247
    },
    {
      "epoch": 2.2142857142857144,
      "grad_norm": 0.12343828025142765,
      "learning_rate": 1.5622042135867838e-05,
      "loss": 0.4189,
      "step": 248
    },
    {
      "epoch": 2.2232142857142856,
      "grad_norm": 0.1157421817524145,
      "learning_rate": 1.5293468531461063e-05,
      "loss": 0.4232,
      "step": 249
    },
    {
      "epoch": 2.232142857142857,
      "grad_norm": 0.10768701281522594,
      "learning_rate": 1.4967568510370498e-05,
      "loss": 0.4259,
      "step": 250
    },
    {
      "epoch": 2.2410714285714284,
      "grad_norm": 0.11855236951996624,
      "learning_rate": 1.464437733941861e-05,
      "loss": 0.4181,
      "step": 251
    },
    {
      "epoch": 2.25,
      "grad_norm": 0.09981025907626229,
      "learning_rate": 1.4323929992293378e-05,
      "loss": 0.4196,
      "step": 252
    },
    {
      "epoch": 2.2589285714285716,
      "grad_norm": 0.11114796288657297,
      "learning_rate": 1.4006261145763657e-05,
      "loss": 0.4218,
      "step": 253
    },
    {
      "epoch": 2.267857142857143,
      "grad_norm": 0.09998416016387277,
      "learning_rate": 1.3691405175926705e-05,
      "loss": 0.4179,
      "step": 254
    },
    {
      "epoch": 2.2767857142857144,
      "grad_norm": 0.11145317342624324,
      "learning_rate": 1.3379396154488142e-05,
      "loss": 0.426,
      "step": 255
    },
    {
      "epoch": 2.2857142857142856,
      "grad_norm": 0.10809564376234657,
      "learning_rate": 1.3070267845075026e-05,
      "loss": 0.4154,
      "step": 256
    },
    {
      "epoch": 2.294642857142857,
      "grad_norm": 0.09865265999187162,
      "learning_rate": 1.2764053699582145e-05,
      "loss": 0.4154,
      "step": 257
    },
    {
      "epoch": 2.3035714285714284,
      "grad_norm": 0.10356277308081305,
      "learning_rate": 1.2460786854552e-05,
      "loss": 0.4249,
      "step": 258
    },
    {
      "epoch": 2.3125,
      "grad_norm": 0.10554928423562794,
      "learning_rate": 1.2160500127589057e-05,
      "loss": 0.4215,
      "step": 259
    },
    {
      "epoch": 2.3214285714285716,
      "grad_norm": 0.10007716549085707,
      "learning_rate": 1.1863226013808392e-05,
      "loss": 0.4221,
      "step": 260
    },
    {
      "epoch": 2.330357142857143,
      "grad_norm": 0.11646959225366664,
      "learning_rate": 1.1568996682319304e-05,
      "loss": 0.4197,
      "step": 261
    },
    {
      "epoch": 2.3392857142857144,
      "grad_norm": 0.10158651685471926,
      "learning_rate": 1.1277843972744127e-05,
      "loss": 0.4135,
      "step": 262
    },
    {
      "epoch": 2.3482142857142856,
      "grad_norm": 0.10217204948795446,
      "learning_rate": 1.0989799391772835e-05,
      "loss": 0.413,
      "step": 263
    },
    {
      "epoch": 2.357142857142857,
      "grad_norm": 0.1080995691928109,
      "learning_rate": 1.070489410975351e-05,
      "loss": 0.4211,
      "step": 264
    },
    {
      "epoch": 2.3660714285714284,
      "grad_norm": 0.08877756673000493,
      "learning_rate": 1.0423158957319331e-05,
      "loss": 0.4199,
      "step": 265
    },
    {
      "epoch": 2.375,
      "grad_norm": 0.0962041869064699,
      "learning_rate": 1.014462442205225e-05,
      "loss": 0.4212,
      "step": 266
    },
    {
      "epoch": 2.3839285714285716,
      "grad_norm": 0.1045928238806768,
      "learning_rate": 9.869320645183861e-06,
      "loss": 0.416,
      "step": 267
    },
    {
      "epoch": 2.392857142857143,
      "grad_norm": 0.09145965237465084,
      "learning_rate": 9.597277418333673e-06,
      "loss": 0.4174,
      "step": 268
    },
    {
      "epoch": 2.4017857142857144,
      "grad_norm": 0.0946988755325339,
      "learning_rate": 9.328524180285226e-06,
      "loss": 0.423,
      "step": 269
    },
    {
      "epoch": 2.4107142857142856,
      "grad_norm": 0.10152439460621095,
      "learning_rate": 9.063090013800488e-06,
      "loss": 0.4247,
      "step": 270
    },
    {
      "epoch": 2.419642857142857,
      "grad_norm": 0.09916208935998831,
      "learning_rate": 8.80100364247265e-06,
      "loss": 0.4201,
      "step": 271
    },
    {
      "epoch": 2.4285714285714284,
      "grad_norm": 0.09330957695586287,
      "learning_rate": 8.542293427617853e-06,
      "loss": 0.4174,
      "step": 272
    },
    {
      "epoch": 2.4375,
      "grad_norm": 0.10014401502360695,
      "learning_rate": 8.286987365206114e-06,
      "loss": 0.4177,
      "step": 273
    },
    {
      "epoch": 2.4464285714285716,
      "grad_norm": 0.09946787106111032,
      "learning_rate": 8.035113082831776e-06,
      "loss": 0.4182,
      "step": 274
    },
    {
      "epoch": 2.455357142857143,
      "grad_norm": 0.09531413850190633,
      "learning_rate": 7.78669783672382e-06,
      "loss": 0.4217,
      "step": 275
    },
    {
      "epoch": 2.4642857142857144,
      "grad_norm": 0.09377218198215026,
      "learning_rate": 7.5417685087963855e-06,
      "loss": 0.4224,
      "step": 276
    },
    {
      "epoch": 2.4732142857142856,
      "grad_norm": 0.09494424619498867,
      "learning_rate": 7.300351603739759e-06,
      "loss": 0.422,
      "step": 277
    },
    {
      "epoch": 2.482142857142857,
      "grad_norm": 0.09167604547691827,
      "learning_rate": 7.062473246152231e-06,
      "loss": 0.4182,
      "step": 278
    },
    {
      "epoch": 2.4910714285714284,
      "grad_norm": 0.09414104830978579,
      "learning_rate": 6.828159177713036e-06,
      "loss": 0.4185,
      "step": 279
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.09349403020134545,
      "learning_rate": 6.597434754396763e-06,
      "loss": 0.4233,
      "step": 280
    },
    {
      "epoch": 2.508928571428571,
      "grad_norm": 0.0893604755299499,
      "learning_rate": 6.370324943729498e-06,
      "loss": 0.4205,
      "step": 281
    },
    {
      "epoch": 2.517857142857143,
      "grad_norm": 0.08281221144451797,
      "learning_rate": 6.146854322086984e-06,
      "loss": 0.4224,
      "step": 282
    },
    {
      "epoch": 2.5267857142857144,
      "grad_norm": 0.08578488209234744,
      "learning_rate": 5.9270470720351305e-06,
      "loss": 0.4133,
      "step": 283
    },
    {
      "epoch": 2.5357142857142856,
      "grad_norm": 0.09120482056235919,
      "learning_rate": 5.7109269797131335e-06,
      "loss": 0.4192,
      "step": 284
    },
    {
      "epoch": 2.544642857142857,
      "grad_norm": 0.0950792789717829,
      "learning_rate": 5.498517432259487e-06,
      "loss": 0.4196,
      "step": 285
    },
    {
      "epoch": 2.553571428571429,
      "grad_norm": 0.09565727805504355,
      "learning_rate": 5.2898414152811715e-06,
      "loss": 0.4223,
      "step": 286
    },
    {
      "epoch": 2.5625,
      "grad_norm": 0.08845162118376791,
      "learning_rate": 5.084921510366321e-06,
      "loss": 0.4207,
      "step": 287
    },
    {
      "epoch": 2.571428571428571,
      "grad_norm": 0.08997378993572731,
      "learning_rate": 4.883779892640532e-06,
      "loss": 0.4235,
      "step": 288
    },
    {
      "epoch": 2.580357142857143,
      "grad_norm": 0.08520848023146284,
      "learning_rate": 4.686438328367287e-06,
      "loss": 0.4169,
      "step": 289
    },
    {
      "epoch": 2.5892857142857144,
      "grad_norm": 0.08661793332561409,
      "learning_rate": 4.492918172592489e-06,
      "loss": 0.4234,
      "step": 290
    },
    {
      "epoch": 2.5982142857142856,
      "grad_norm": 0.08401595957495671,
      "learning_rate": 4.303240366833596e-06,
      "loss": 0.4139,
      "step": 291
    },
    {
      "epoch": 2.607142857142857,
      "grad_norm": 0.08238451078646887,
      "learning_rate": 4.117425436813438e-06,
      "loss": 0.4202,
      "step": 292
    },
    {
      "epoch": 2.616071428571429,
      "grad_norm": 0.08738609784414623,
      "learning_rate": 3.935493490239046e-06,
      "loss": 0.4208,
      "step": 293
    },
    {
      "epoch": 2.625,
      "grad_norm": 0.08259864478704049,
      "learning_rate": 3.7574642146257634e-06,
      "loss": 0.4186,
      "step": 294
    },
    {
      "epoch": 2.633928571428571,
      "grad_norm": 0.08659177417529826,
      "learning_rate": 3.583356875166737e-06,
      "loss": 0.4237,
      "step": 295
    },
    {
      "epoch": 2.642857142857143,
      "grad_norm": 0.08119845593756564,
      "learning_rate": 3.4131903126482045e-06,
      "loss": 0.4154,
      "step": 296
    },
    {
      "epoch": 2.6517857142857144,
      "grad_norm": 0.08611600596660274,
      "learning_rate": 3.2469829414106503e-06,
      "loss": 0.4201,
      "step": 297
    },
    {
      "epoch": 2.6607142857142856,
      "grad_norm": 0.08276786453261639,
      "learning_rate": 3.0847527473561167e-06,
      "loss": 0.4248,
      "step": 298
    },
    {
      "epoch": 2.669642857142857,
      "grad_norm": 0.08240160307178532,
      "learning_rate": 2.9265172860018754e-06,
      "loss": 0.415,
      "step": 299
    },
    {
      "epoch": 2.678571428571429,
      "grad_norm": 0.08006883576509521,
      "learning_rate": 2.7722936805807267e-06,
      "loss": 0.4171,
      "step": 300
    },
    {
      "epoch": 2.6875,
      "grad_norm": 0.08381933167379557,
      "learning_rate": 2.622098620187963e-06,
      "loss": 0.427,
      "step": 301
    },
    {
      "epoch": 2.696428571428571,
      "grad_norm": 0.08303705152303609,
      "learning_rate": 2.475948357975435e-06,
      "loss": 0.4198,
      "step": 302
    },
    {
      "epoch": 2.705357142857143,
      "grad_norm": 0.08870666410228789,
      "learning_rate": 2.3338587093927245e-06,
      "loss": 0.4219,
      "step": 303
    },
    {
      "epoch": 2.7142857142857144,
      "grad_norm": 0.0818167400851114,
      "learning_rate": 2.195845050475698e-06,
      "loss": 0.4223,
      "step": 304
    },
    {
      "epoch": 2.7232142857142856,
      "grad_norm": 0.08139799765060797,
      "learning_rate": 2.0619223161825942e-06,
      "loss": 0.4184,
      "step": 305
    },
    {
      "epoch": 2.732142857142857,
      "grad_norm": 0.07642422600278949,
      "learning_rate": 1.932104998777877e-06,
      "loss": 0.4179,
      "step": 306
    },
    {
      "epoch": 2.741071428571429,
      "grad_norm": 0.07996683628078224,
      "learning_rate": 1.8064071462639798e-06,
      "loss": 0.4184,
      "step": 307
    },
    {
      "epoch": 2.75,
      "grad_norm": 0.07773519961892873,
      "learning_rate": 1.684842360861092e-06,
      "loss": 0.4261,
      "step": 308
    },
    {
      "epoch": 2.758928571428571,
      "grad_norm": 0.07860341250434275,
      "learning_rate": 1.5674237975352457e-06,
      "loss": 0.4171,
      "step": 309
    },
    {
      "epoch": 2.767857142857143,
      "grad_norm": 0.0784236768243464,
      "learning_rate": 1.4541641625747428e-06,
      "loss": 0.4192,
      "step": 310
    },
    {
      "epoch": 2.7767857142857144,
      "grad_norm": 0.0761682477040851,
      "learning_rate": 1.3450757122151826e-06,
      "loss": 0.4129,
      "step": 311
    },
    {
      "epoch": 2.7857142857142856,
      "grad_norm": 0.07767572155515515,
      "learning_rate": 1.2401702513131597e-06,
      "loss": 0.4182,
      "step": 312
    },
    {
      "epoch": 2.794642857142857,
      "grad_norm": 0.07941885760590452,
      "learning_rate": 1.1394591320688098e-06,
      "loss": 0.4186,
      "step": 313
    },
    {
      "epoch": 2.803571428571429,
      "grad_norm": 0.07998708863714857,
      "learning_rate": 1.0429532527973695e-06,
      "loss": 0.4263,
      "step": 314
    },
    {
      "epoch": 2.8125,
      "grad_norm": 0.0783378129960652,
      "learning_rate": 9.50663056749801e-07,
      "loss": 0.4141,
      "step": 315
    },
    {
      "epoch": 2.821428571428571,
      "grad_norm": 0.07670652617883925,
      "learning_rate": 8.625985309827211e-07,
      "loss": 0.4192,
      "step": 316
    },
    {
      "epoch": 2.830357142857143,
      "grad_norm": 0.07288684797979382,
      "learning_rate": 7.787692052776363e-07,
      "loss": 0.409,
      "step": 317
    },
    {
      "epoch": 2.8392857142857144,
      "grad_norm": 0.07510033336815654,
      "learning_rate": 6.991841511097175e-07,
      "loss": 0.4167,
      "step": 318
    },
    {
      "epoch": 2.8482142857142856,
      "grad_norm": 0.07666734447787482,
      "learning_rate": 6.238519806661192e-07,
      "loss": 0.4191,
      "step": 319
    },
    {
      "epoch": 2.857142857142857,
      "grad_norm": 0.07597379543452941,
      "learning_rate": 5.527808459140449e-07,
      "loss": 0.4196,
      "step": 320
    },
    {
      "epoch": 2.866071428571429,
      "grad_norm": 0.07587470302303759,
      "learning_rate": 4.859784377185639e-07,
      "loss": 0.4188,
      "step": 321
    },
    {
      "epoch": 2.875,
      "grad_norm": 0.07535633004319045,
      "learning_rate": 4.2345198501038796e-07,
      "loss": 0.4138,
      "step": 322
    },
    {
      "epoch": 2.883928571428571,
      "grad_norm": 0.0777275889451964,
      "learning_rate": 3.652082540035906e-07,
      "loss": 0.4231,
      "step": 323
    },
    {
      "epoch": 2.892857142857143,
      "grad_norm": 0.07632262719194602,
      "learning_rate": 3.112535474633971e-07,
      "loss": 0.4214,
      "step": 324
    },
    {
      "epoch": 2.9017857142857144,
      "grad_norm": 0.07651730781426783,
      "learning_rate": 2.6159370402415674e-07,
      "loss": 0.4186,
      "step": 325
    },
    {
      "epoch": 2.9107142857142856,
      "grad_norm": 0.07477013510660417,
      "learning_rate": 2.1623409755751056e-07,
      "loss": 0.4248,
      "step": 326
    },
    {
      "epoch": 2.919642857142857,
      "grad_norm": 0.07897064466517018,
      "learning_rate": 1.751796365908831e-07,
      "loss": 0.4233,
      "step": 327
    },
    {
      "epoch": 2.928571428571429,
      "grad_norm": 0.07717251326428437,
      "learning_rate": 1.3843476377628507e-07,
      "loss": 0.4177,
      "step": 328
    },
    {
      "epoch": 2.9375,
      "grad_norm": 0.07474269162329465,
      "learning_rate": 1.0600345540958234e-07,
      "loss": 0.4185,
      "step": 329
    },
    {
      "epoch": 2.946428571428571,
      "grad_norm": 0.07632913690300189,
      "learning_rate": 7.788922100020024e-08,
      "loss": 0.4227,
      "step": 330
    },
    {
      "epoch": 2.955357142857143,
      "grad_norm": 0.07484327906269052,
      "learning_rate": 5.4095102891338344e-08,
      "loss": 0.4296,
      "step": 331
    },
    {
      "epoch": 2.9642857142857144,
      "grad_norm": 0.07247458214414504,
      "learning_rate": 3.462367593075833e-08,
      "loss": 0.4137,
      "step": 332
    },
    {
      "epoch": 2.9732142857142856,
      "grad_norm": 0.07522031543191121,
      "learning_rate": 1.9477047192140164e-08,
      "loss": 0.4274,
      "step": 333
    },
    {
      "epoch": 2.982142857142857,
      "grad_norm": 0.07361261056246672,
      "learning_rate": 8.656855747082305e-09,
      "loss": 0.4179,
      "step": 334
    },
    {
      "epoch": 2.991071428571429,
      "grad_norm": 0.07312212144143204,
      "learning_rate": 2.1642724877146693e-09,
      "loss": 0.411,
      "step": 335
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.07511641768667605,
      "learning_rate": 0.0,
      "loss": 0.4162,
      "step": 336
    },
    {
      "epoch": 3.0,
      "step": 336,
      "total_flos": 8.070235818242867e+18,
      "train_loss": 0.4864790922119504,
      "train_runtime": 20111.4838,
      "train_samples_per_second": 8.547,
      "train_steps_per_second": 0.017
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 336,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 8.070235818242867e+18,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}